A DNA ALGORITHM FOR THE BATIMETRIC MAPPING IN THE LAGOON OF 
VENICE USING QUICKBIRD MULTISPECTRAL DATA 
M. Gianinetto?, G. Lechi ^ *, 
? DIIAR, Politecnico di Milano, P.zza Leonardo da Vinci 32, 20133 Milano, Italy 
marco.gianinetto@polimi.it 
giovanmaria.lechi@polimi.it 
KEY WORDS: Remote Sensing, Bathymetry, Mapping, High resolution, Multispectral, Quickbird. 
ABSTRACT: 
During the last decade, several studies have focused on the use of passive multispectral remote sensing to derive the bathymetry in 
coastal zone. In particular, data acquired with the SPOT and the Landsat TM/ETM+ sensors have been used to derive models of 
bathymetry at medium scales. Until now, the successful application of passive remote sensing techniques to bathymetry mapping 
was restricted to costal zones with clear water and small changes in the seabed, but with the availability of the high resolution 
satellites (IKONOS, Eros-Al1, QuickBird, SPOT-5), researchers have a new powerful tools to study environmental phenomenon at 
large scale. This paper focus on the use of high resolution imagery to estimate water depths in a lagoon environment. Starting from 
the depth of penetration zone method proposed by Jupp for costal bathymetry mapping, a new genetic algorithm was developed for 
lagoon bathymetry mapping. The potential use of the QuickBird multispectral data, together with the new algorithm developed, was 
tested in a complex environment such as the lagoon of Venice (Italy). Several tests have been performed into five different test sites 
(S.Erasmo littoral, Treporti canal, S. Felice canal, Canesa canal and Bari canal), where 18 radiometric transects were traced to study 
the lagoon bathymetry. The accuracy of the batimetric measures was assessed by using other known soundings depth points within 
the test area. An interesting correlation between the real and the computed bathymetry was found. The limit of a such analysis lies 
in the correct calibration of the model, that, for the complex lagoon ecosystem, is not a simple task. 
1. INTRODUCTION 
The bathymetry is that branch of the oceanography that deals 
with the measure of the depths of the sea and studies its 
configuration and evolution over the time. Nowadays, 
bathymetric surveys are generally carried out using ships 
equipped with sonar instruments or, in the case of coastal zones, 
by Airborne Laser Bathymetry (ALB) from aerial platform. The 
implementation, but especially the updating, of the bathymetric 
maps using the traditional techniques shows high costs and long 
time to realize the survey and process all the data collected 
(Abduallah ef al., 1991), for this reason researches have become 
interested in use of remote sensing (RS) techniques to find a 
fast and low-cost mapping tool. 
In the course of the last decade, several studies have shown the 
possibility to use the multispectral data acquired with the SPOT 
and Landsat satellites to derive the bathymetry in coastal zone 
(Jupp, 1988; Nordman ef al., 1990; Hengel and Spitzer, 1991), 
but frequently with precisions not suitable to the editing of 
charting for the navigation. Until now, the successful 
application of passive RS techniques to bathymetry mapping 
was restricted to shallow and clear water, with small changes in 
the seabed (Gordon and McCluney, 1975; Lyzenga, 1986; 
Bierwrth et al. 1992). 
This paper focus on the use of satellite’s high resolution images 
(HRI) to estimate water depths in a lagoon environment. The 
small pixel size of such sensors («5 m), in respect to Landsat 
TM/ETM- (30 m) and SPOT HRV (20 m), could be a key 
factor to obtain accurate batimetric computations. In particular, 
the potential use of the QuickBird multispectral data was tested 
  
* Corresponding author. 
94 
in the lagoon of Venice (Italy) using a new algorithm developed 
(Gianinetto et a/., 2003). 
2. BATHYMETRY MAPPING ALGORITHM 
2.1 Jupp DOP model 
Electromagnetic radiation that cross a water column suffer an 
attenuation due to the joint action of diffusion and absorption of 
the medium. Considering a bundle of monochromatic light, the 
relative loss of radiant flux is proportional to the size of the 
path, to less of a coefficient of proportionality (extinction 
coefficient). A model that finds large employment in literature 
to reconstruct the bathymetry in coastal zones from RS 
multispectral data, is the depth of penetration zone (DOP) 
method proposed by Jupp (Jupp, 1988) 
Le = 2k, + ( - ea». (1) 
L. = measured at-sensor radiance 
L, = emergent radiance from the seabed 
Ly = emergent radiance from the different layers 
z = depth 
k = coefficient of absorption 
Hypothesizing negligible the term of L,, directly related to 
quality of the water (suspended sediments), and for small 
changes in seabed, then, among the depth of the water column 
and the logarithm of the measured at-sensor radiance there is a 
linear correspondence. Under this conditions, rearranging Eq. 
International A 
AIC UO 
(1) lead to th 
determination 
where N=1 
In practice, to 
constant coeff 
later, this is th 
in the lagoon « 
is very high. 
22 A new 
measures in tl 
The high spat 
suspended sed 
make inapplic: 
high variabilit 
in the lagoor 
homogeneity i 
less one of t 
mapping mod 
seabed in the 
has been mod 
of the real bat 
lagoon of Ven 
is interpretat 
geometric me 
perform a co 
but uses only 
absorption (tk 
and, subseqi 
separately th 
model, called 
has been elim 
been introdu 
multispectral 
where m- 
The new alg 
genetic algor 
electromagne 
intervals of i 
water columr 
for each of t 
the paramete